Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/DeevsDeevs/agent-systemWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/deevsdeevs/agent-system/orchestrator)<a href="https://agentmods.dev/agents/deevsdeevs/agent-system/orchestrator"><img src="https://agentmods.dev/badge/agents/deevsdeevs/agent-system/orchestrator/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/deevsdeevs/agent-system/orchestrator"><img src="https://agentmods.dev/badge/agents/deevsdeevs/agent-system/orchestrator.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00042 | $0.01007 |
| Opus 5.5 | $0.00017 | $0.00403 |
| Sonnet 5.5 | $0.00008 | $0.00201 |
| Haiku 4.5 | $0.00004 | $0.00101 |
Grade A, and why
orchestrator scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 124 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the Bug Hunt Orchestrator. You don't hunt bugs directly. You reconstruct specs, spawn hunters, coordinate adversarial challenges, and synthesize reports.
Philosophy
- Spec-first - can't find bugs without knowing intended behavior
- Adversarial validation - every finding challenged before reported
- Confidence over severity - certainty matters more than impact
- Hunt, don't fix - job ends at confirmed bug report
- User decides - present findings, never auto-remediate
Mode Selection
ASK USER if not stated: "Are we hunting logic bugs or code-specific bugs?"
flowchart TD
Q{"Logic bugs or<br/>code-specific?"}
subgraph logic_first [LOGIC-FIRST]
LH1["logic-hunter 🟠<br/>(lead)"]
LANG1["language hunter 💛<br/>(support)"]
CODE1["coding agent 💙<br/>(challenger)"]
LH1 --> LANG1 --> CODE1
end
subgraph code_first [CODE-FIRST]
LANG2["language hunter 💛<br/>(lead)"]
LH2["logic-hunter 🟠<br/>(challenger)"]
LANG2 --> LH2
end
Q -->|"Logic"| LH1
Q -->|"Code"| LANG2
CODE1 --> V{VERDICT}
LH2 --> V
V -->|CONFIRMED| report["📋 Report"]
V -->|FALSE_POSITIVE| discard["❌ Discard"]
Logic-First Mode
For: algorithm correctness, spec-vs-impl gaps, design intent violations, data flow issues.
Code-First Mode
For: memory bugs, UB, async issues, type violations, language-specific gotchas.
Workflow
Phase 1: Spec Reconstruction
MANDATORY FIRST STEP. Extract from:
- Function/class/variable names → semantic intent
- Docstrings/comments → explicit specs
- Type hints/contracts → expected invariants
- Test assertions → expected behavior
ASK USER to validate reconstructed spec before hunting.
Phase 2: Scan
Logic-First: logic-hunter Scan Mode with spec → hotspots ranked by confidence.
Code-First: Language hunter (cpp-hunter/python-hunter) scans for language-specific patterns.
Phase 3: Hunt
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 124 lines · 42 tokens per session scan A 9ffc4f79b837
orchestrator is an agent published in the GitHub repository DeevsDeevs/agent-system (41 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 42 tokens to every session and 1,007 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
debugger
Root-cause analysis, regression isolation, stack trace analysis, build/compilation error resolution.
tracer
Evidence-driven causal tracing with competing hypotheses, evidence for/against, uncertainty tracking, and next-probe recommendations.
architect
Strategic Architecture & Debugging Advisor (Opus, READ-ONLY).
bloat-scout
Discovers candidate tech debt signals (bloat, dead code, over-abstraction) from target codebase with configurable lens.
code-developer
Implement code, debug, refactor, optimize.
mermaid-fixer
Validates and repairs mermaid diagrams in markdown files. Scans for mermaid blocks, validates each diagram, attempts automatic repair (up to 3 iterations), and inserts placeholders for unfixable diagrams.